Bayesian techniques and the principle of maximum entropy in ion-beam analysis applications

The Bayesian Formalism (BF) and the principle of Maximum Entropy (MaxEnt) are applied to techniques relevant to Ion-Beam Analysis (IBA). The BF is an efficient and theoretically sound method of information recovery, and when applied together with an entropic prior it has special applications in solving ill-posed inverse problems, e.g. the deconvolution of the detector response function from PIXE or RES spectra. Results include the recovery of depth profiles in PIXE using a two-detector set-up. The deconvolution of RES spectra using the BF is also compared to more conventional deconvolution techniques, namely Fourier Transforms and the non-linear Jansson method. (C) 1998 Published by Elsevier Science B.V.